Tuberculosis is one of the greatest infectious causes of disease and mortality worldwide. The problem has been exacerbated in recent years by the emergence of drug resistant strains and the concurrent HIV epidemic. There is no reliable vaccine against tuberculosis and there is increasing evidence that natural infection and disease cannot prevent subsequent reinfection. Thus host immunity is only partially effective in preventing and eliminating infection. To aid in the understanding of the basic pathogenesis of tuberculosis in the context of host immunity, we will use a closely related pathogenic species Mycobacterium marinum, the agent of a systemic tuberculosis-like disease in ectotherms, and a peripheral granulomatous disease in humans. M marinum has become established as a powerful model for tuberculosis owing to its genetic and pathogenic similarities to M. tuberculosis, its relatively rapid growth, and the existence of multiple natural hosts that serve as animal models. We use adult leopard frogs and zebrafish to study pathogenesis as well as zebrafish embryos, which allow real-time monitoring of host-pathogen interactions due to their optical transparency. Using the embryo model in conjunction with M. marinum granuloma-activated genes that we previously identified, we have been able to monitor granuloma formation in real-time. We will continue to use all of these models to determine how mycobacteria persist in the face of a robust host immune response in vivo. We will use the zebrafish embryo model to determine the cellular mechanism of attenuation of known Mycobacterium virulence genes, and in so doing, develop an integrated understanding of the events leading to chronic infection. As a first step, we have already determined that the RD1 virulence locus specifically promotes the aggregation of macrophages in vivo. We will study the in vivo trafficking of wild type and mutant strains of M marinum in the context of previously formed granulomas, with special attention paid to the role of the caseum. Finally, we will perform experiments to directly determine the physiological and replicative state of mycobacteria located within chronic granulomas. These experiments will determine whether such bacteria are in a quiescent state or are actively replicating. The information obtained from our proposed experiments will further our fundamental knowledge about host-pathogen interactions in tuberculosis.